The four peptide segments or motifs which form the active site position the conserved Asp nucleophile, Asp acid/base, the Lys/Arg and Ser/Thr phosphate-binding residues and the Mg^* cofactor Asp/Glu binding residues (Figure IB). These residues, in combination with the scaffold main-chain elements, form a steric and electrostatic mold that stabilizes the trigonal bipyramidal transition states/intermediates produced along the reaction pathway (Figure 1B) [7]. The HAD phosphatase substrate recognition elements are located in either a cap domain (as in HAD classes C1 and C2, also known as Type I and Type 11) tethered to the core domain by a solvated linker, or in short loop/helical segments that extend from the core domain (as in the capless HAD class CO also known as Type III) (Figure 1A) [8]. Although HAD phosphatases possess the same catalytic site and proceed through the same second partial reaction, they are able to use the specific structural requirements of the substrate-binding step and the subsequent addition-elimination steps of the first partial reaction to discriminate between the physiological substrate and other phosphorylated species (macromolecules and metabolites). The induced fit model, wherein substrate binding is followed by cap domain or loop closure, applies to most HAD phosphatases. Favorable electrostatic interaction between the substrate leaving group and the cap domain/gating loops will contribute to the substrate-binding affinity. For efficient turnover, the phosphoryl group must be bound in the correct orientation within the catalytic site. If the substrate-leaving group is too large or too small, nonproductive binding is likely to occur. Thus, the size, shape and electrostatic surface ofthe active site region that extends from the catalytic site to the active site entrance can provide significant insight into the identity of the physiological substrate. This serves as the basis for the use of virtual screening (made possible by the Structure Core and Computation Core) to identify candidates for the physiological substrate herein. Substrate specificities defined by experimental activity screens suggest that the typical HAD phosphatase has loose substrate specificity coupled with modest catalytic efficiency. Thus, activity screens alone often cannot identify the actual physiological substrate. Rather, they provide candidates that can be further interrogated using the tools provided by the Sequence/Genome Analysis Core and Microbiology Core.